Interactive Effects of Climate Change, Wetlands, and Dissolved Organic Matter on UV Damage to Aquatic Foodwebs U.S. Environmental Protection Agency’s Global Change and Ecosystem Protection Research STAR Progress Review Workshop June 16-18, 2004

Principal Investigators: Scott Bridgham 1, Gary Lamberti 2, David Lodge 2, Patricia Maurice 2, Carol Johnston 3, Boris Shmagin 3 Postdoctoral Associate: Paul Frost 2 Ph.D. Students: James Larson 2, Kathryn Young 2, Zhiyu Zheng 3 Research Technician: Christine Cherrier 1 1 Univ. of Oregon, 2 Univ. of Notre Dame, 3 South Dakota State Univ.

Overarching Goal Provide a better understanding of how land use, climate, and UVR affect foodweb structure in streams and rivers through their complex interactions with DOM, landscape characteristics, and climate.

Stream UVR-DOM Study (SUDS)

Five Main Objectives 1.Relate DOM concentration and chemical characteristics to discharge, landscape characteristics, and stream geomorphology. 2.Determine how in-stream processing of DOM through biodegradation and photodegradation varies spatially within the watershed. 3.Determine how various climate change scenarios will affect discharge and, thus, DOM concentration at a variety of spatial scales.

Objectives, Cont. 4.Determine interactions among UVR intensity and DOM concentration and chemistry. 5.Determine the response of stream foodwebs to the interactions among UVR intensity and DOM concentration and type.

Objectives 1.Relate DOM concentration and chemical characteristics to discharge, landscape characteristics, and stream geomorphology.

Study Sites Ontonagon watershed km 2 watershed -drains into Lake Superior -streams 1 st to 6 th order -60 sampling sites in Sept sites sampled ~ 2 months for 2 years

FactorMeanMin.Max % of area in wetland % of area in lake % of area in agriculture watershed area (km 2 ) total stream length (km) drainage density (km km -2 ) Characteristics of Ontonagon sub-watersheds

Sept Sampling

Why the wide range in DOC among these streams? % lake % developed % evergreen % agriculture % wetland log stream length log watershed area log watershed perimeter log drainage density log maximum slope log mean slope log standard deviation slope stream geomorphology landscape features

ln DOC (mg C L -1 )

Multiple factor regression [DOC] Predictor Variables (5) % lake (-) % developed % evergreen % agriculture (4) % wetland (+) log stream length (2) log watershed area (-) log watershed perimeter (1) log drainage density (-) log maximum slope (3) log mean slope (-) log standard deviation slope p< r 2 = 0.61

DOC Molecular Weight Predictor Variables (2) % lake (-) % developed % evergreen % agriculture % wetland log stream length log watershed area log watershed perimeter (1) log drainage density (-) log maximum slope log mean slope log standard deviation slope p< , r 2 = 0.54

Sept. 2002

May + August 2003 DOM Data No Lakes Lake Outflows DOC mg/L (95% Confidence) Average [DOC] from Combined May and August 2003 Absorbance at 280 nm divided by the moles C per liter Abs280 * L/mol C No LakesLake Outflows

Ongoing Landscape DOM Projects  Expand GIS database by adding surficial geology, soil type, and soil C:N ratio. Compare different wetland databases and determine if wetland type is an important variable.  Examine how landscape relationships with DOM concentration and chemistry vary with seasonally with ~ bimonthly sampling of stream survey.

Ongoing Landscape DOM Projects, Cont.  Explore how DOM concentration and chemistry vary longitudinally in streams with and without lake outlets.

Objectives 2.Determine how in-stream processing of DOM through biodegradation and photodegradation varies spatially within the watershed.

Results to Date  Biodegradation of high molecular-weight DOM is faster, and the low-molecular weight fraction is preferentially degraded.  Biodegradation rates of DOM are dependent on microbial community structure.

Ongoing DOM Experiments  Examine short- and long-term photodegradation and biodegradation rates of DOM from six different stream sources.  With and without nutrient addition.  How important is prior photodegradation in biodegradation?

Objectives 3.Determine how various climate change scenarios will affect discharge and, thus, DOM concentration at a variety of spatial scales.

 Factor analysis has been used at the scale of the conterminous U.S., the Great Lakes region, and the Upper Great Lakes region to determine landscape and climatic correlates of annual and seasonal discharge in streams and rivers.

 We will gather as many [DOC] vs. discharge data as can be found for the Upper Great Lakes. Various climate change scenarios will be applied to these models to predict effects on DOM concentrations and UVR penetration into the water column.  Mechanistic hydrological model for the Ontonagon Watershed?

Objectives 4.Determine interactions among UVR intensity and DOM concentration and chemistry.

k d = - ln (I d /I o )/ depth The “K d ” tells you how fast light disappears in the water Big k d  Little light Small k d  Lots of light

UVB (μW cm -2 ) Stream UVB Model irradiance benthos = irradiance top x canopy attenuation x water attenuation

Ongoing UVR Landscape Research  Mapping UVR penetration within the entire Ontonagon Watershed.  Quantifying UVR dose spatially within a number of streams with ‘dosimetry’ strips.

Objectives 5.Determine the response of stream foodwebs to the interactions among UVR intensity and DOM concentration and type.

Controlled experiments to examine the interactive effects of UVR and DOM on stream food web structure

Experiment: Change UV flux onto periphyton by altering DOM concentration and through the use of plastic UV screens w/ plasticno plastic plus DOCno UVB high DOC low UVB high DOC no DOCno UVB low DOC high UVB low DOC 4 replicates per treatment combination

chlorophyll (  g cm -2 )

particulate carbon (  g cm -2 )

NO3 (  g L -1 ) SRP (  g L -1 ) NH4 (  g L -1 )

Ongoing Foodweb Experiments: How does light intensity interact with UVR to affect periphyton growth? + shading (-90%) no shading ambient - UVB -UVA + -UVB

Ongoing Foodweb Experiments: How do nutrients and DOM molecular weight interact to affect foodwebs? + N, +P- N, - P no DOM (ground water) + HMW DOM (~ 8 mg/L) + LMW DOM (~ 8 mg/L) Periphyton; last week snails, mayflies, caddisflies, chironomids, amphipods UVR excluded in all treatments